RNNs · Level 3 of 5
Bidirectional RNN
A recurrent design processing a sequence in both directions.
It needs future context and therefore is not directly causal for live prediction.
Example
A completed sentence is encoded from left and right.
Listen to the definition and example
Audio transcript
Bidirectional RNN. A recurrent design processing a sequence in both directions. It needs future context and therefore is not directly causal for live prediction. For example: A completed sentence is encoded from left and right.
Explore this concept
Why it matters
This helps you trace information and gradient flow through sequence models.
Start with
Quick recall question
Try answering before looking back at the definition.